Cycle Theory
Introduction
Developing a useful model of the stock market environment is no mean feat.
------------------------------------------------------------------------------
--------------------------------------------------
The aim of a stock market model is to predict the price of specific stocks or market averages. (Following Richard Ney, the aim might be to learn what the specialists are doing.) A model helps you build a strategy to play the market successfully. Various approaches have been adopted to try to predict prices.
Price charts of companies -- particularly long-term charts, which aren't easy to manipulate -- show that price tends to correlate somewhat with the earnings growth of the companies, so long as the earnings growth is steady and smooth. The greater and longer the earnings growth, the higher the price. I.e., prices long-term tend to move with market fundamentals -- with how well companies and the overall economy are doing. Continued growth tends to build confidence in future growth and price increase.
There is little doubt, however, that differences exist between price movement and earnings growth, at least in the shorter term. In other words, price movement isn't always in phase with earnings growth. In fact, prices oscillate even when earnings rise steadily. This is true even when you consider the dividends paid out by the company, considered by some to be a better indicator of price movement. Apparently, other factors than purely fundamental ones are involved. It isn't entirely clear what these factors are, but they seem to depend on the market psychology, on investor expectations -- how investors feel about the market -- how bullish or bearish they are.
J. W. Hurst has characterized this peculiarity as the X Motivation factor, the factor that produces the oscillatory motion. He says that 23% of all price motion is oscillatory in nature and semi-predictable. This cyclical portion of the sum of the influences provides the basis for market trends and thus for market timing. Inasmuch as cycles are the determining factors for the prediction of prices, it behooves us to understand something about the beasts.
----------------------------------------------
We have cycles of day and night. Our eating habits display the cycle of meals. There are weekly, monthly, and yearly cycles. Elections take place in two-, four-, and 6-year cycles. And the markets exhibit cyclical behavior. Stock prices move through a sequence that turns back on itself again and again.
The cyclic approach to prediction presumes driving forces that act in cyclical fashion on the stock and that add together to determine its price action. This presumption is a modeling tool that lets us use traditional concepts of cyclicality to elicit information. The cyclicality takes the form of sine and cosine functions, or Trigonometry.
In one direction -- from elements to their sum -- a number of these functions can be combined appropriately (synthesized) to generate actual price patterns. In the other direction, a sum (price pattern) can be analyzed to get its components.
(The former procedure is known as synthesis, and the latter is known as spectral analysis (also harmonic analysis) and typically invokes Fourier Analysis to identify the frequency components whose sum makes up the pattern.)
We identify three essential properties of cyclical components:
------------------------------------------